Explicitly exit at end of execution

[kjrokos]

Resolves #359

Notes from investigation:

• OkHttp has the hanging issue, but the exact same code (as close as Ktor allows, which is quite close) using the Cio engine does not.
• In both cases, the server gets the result back in approximately the same amount of time. The hanging occurs after all other execution is finished.
• In the OkHttp case, the culprit appears to be an "OkHttp Dispatcher" thread that is still running. I've found a good number of issues related to this, but all from around ~2020 and claim be resolved in the versions we are using (e.g., https://youtrack.jetbrains.com/issue/KTOR-396)
• I tried manually closing the client, engine, dispatcher, executor, connection pool, etc - with no success. As near as I can tell, closing just the client should be sufficient, but clearly isn't in our case
• If I run the exact same code outside of a container, I don't have this problem. I don't know what is different about the container environment. It's also possible (maybe?) that there is a thread sticking around longer but it's not noticeable because the system itself isn't waiting on it to close. But I'm not sure how to test that, and it seems unlikely.
• If we explicitly call 'System.exit(0)' at the end of collection, everything works as expected. I don't love it, but this is likely the best (and most general) solution. The user shouldn't be doing anything at this point - any cleanup should have already happened, all results have already been sent back to the server, and the user would expect that the program exits here immediately anyway.
• I don't know if the issue can present itself in Python in the same way, but I have added exits at the same place in the Python templates for consistency